Research Of Logistics Distribution Routing Problem Based On PSO

With the market economy by leaps and bounds and the development of modern logistics technology, logistics and distribution sectors, is being increasingly widespread concern and distribution of logistics and distribution routing problem (VRP) has become the core issue of logistics and distribution. Logistics Distribution Vehicle Routing Optimization is a both theoretical and practical significance and challenging subject, not only because it is an NP-hard problem, but mainly because it has strong economic benefits, so the cause the interest of many researchers. Of VRP belongs to a class of difficult combinatorial optimization problems, the study has important theoretical value and practical significance. It is in this context, focus on a variety of traditional algorithms to be compared with an improved particle swarm optimization algorithm based on neural networks designed on this basis.In this paper, the multiple depot vehicle routing problem in logistics and distribution problems, according to the shortcomings of the standard neural network algorithm designed in this paper the adaptive mutation particle swarm optimization. Maintain the standard PSO algorithm search mechanism under the premise of the particles to the local convergence, the mutation rate of the current particle adaptive adjustment, thus avoiding premature convergence. In addition, for the multiple depot vehicle routing problem, the paper gives a new encoding modes, reducing the probability of infeasible solutions. MatLab2011a platform simulation results proved that this algorithm outperformed the standard PSO algorithm in solving such problems, while maintaining good global search capability, but also effectively prevent prematurity convergence.